# GraphN pricing

Published usage pricing from the billing catalog. Unpublished rates render as unavailable.

## LLM inference (USD per 1M tokens)

- **qwen3-80b**: Input $0.20/1M · Output $0.60/1M
- **qwen3-235b**: Input $0.22/1M · Output $0.88/1M
- **qwen3-coder**: Input $0.90/1M · Output $0.90/1M
- **nemotron-3-super**: Input $0.90/1M · Output $0.90/1M
- **gpt-oss-120b**: Input $0.15/1M · Output $0.60/1M
- **qwen2.5-vl-7b-instruct**: Input $0.05/1M · Output $0.05/1M
- **qwen3-vl**: Input $0.10/1M · Output $0.15/1M
- **qwen3.8-27b**: Input $0.45/1M · Output $3.20/1M
- **qwen3.5-122b-a10b-fp8**: Input $0.29/1M · Output $2.40/1M
- **gemma-4**: Input $0.39/1M · Output $0.97/1M
- **gemma-4-e4b**: Input $0.15/1M · Output $0.60/1M
- **custom:\***: billed by GPU runtime on GraphN-managed GPUs
- **imported:\***: billed by your provider account, not by GraphN

## Custom model GPU runtime

- **Per GPU-hour**: $5.00/GPU-hr
- **Per GPU-second**: $0.00138889/GPU-second

## Platform meters

- **KB embedding tokens**: $0.02/1M input tokens
- **KB rerank tokens**: $0.02/1M input tokens
- **Unstructured RAG ingestion**: $4.00/1,000 pages
- **Markdown conversion**: $0.50/1,000 pages
- **Object storage**: $0.069/GB-month (hourly occupancy)
- **Knowledge base vector storage**: $1.10/GB-month (hourly occupancy)
- **Function invocation**: $0.40/1M calls
- **Connector invocation**: $0.40/1M calls
- **Function compute**: $0.0000166667/GB-second
- **RAG job trigger**: $0/call

- [Managed inference](https://graphn.ai/inference)
- [Create account](https://graphn.ai/signup)